1515(3):
(3):358-xxx
358-365 (2008)
(2008)
Site-specific features affect pollination success
of a gynodioecious understory shrub in a
gender-specific mode1
Conchita ALONSO2 & Carlos M. HERRERA, Estación Biológica de Doñana, Consejo Superior de
Investigaciones Científicas (CSIC), Avda. María Luisa s/n, Pabellón de Perú, 41013, Sevilla, Spain,
e-mail: [email protected]
Abstract: Daphne laureola (Thymelaeaceae) is a gynodioecious, winter-flowering shrub inhabiting the understory of shady
mountain forests. The small, inconspicuous yellowish-green flowers are pollinated by ectothermic insects whose activity is
mainly confined to infrequent spells of sunny weather. We tested the hypotheses that (1) the solar irradiance environment of
individual D. laureola plants and distance to the nearest conspecific should influence their pollination success and (2) the
effect of both site-specific features on pollination success should be sex-dependent, given the difference between female and
hermaphrodite flowers in conspicuousness and floral rewards. We estimated maternal pollination success (mean number of
pollen tubes/pistil and proportion of unfertilized flowers), measured the distance to the nearest flowering conspecific, and
evaluated the irradiance environment (percent of canopy cover) for individual D. laureola plants in 7 southeastern Spanish
populations. On average, hermaphrodite flowers were pollinated more often and had more pollen tubes/pistil than females.
Distance to the nearest conspecific affected individual pollination success negatively, but the effect was significant only for
females. Furthermore, there existed a highly significant sex × cover interaction effect on pollination success. Among female
individuals, the number of pollen tubes/pistil decreased and the proportion of unfertilized flowers increased from closed to
more open sites, while hermaphrodites showed non-significant and opposite trends. The contrasting sign of the relationship
between cover and pollination success for the 2 genders is expected to generate a fine-scale, within-habitat mosaic in the
magnitude of the pollination advantage of hermaphrodite individuals, which would be greater in the more open, sunny locations
where ectothermic pollinators should be more active. Site-specific differences in relative pollination success of sexes might
contribute to the persistence of gynodioecy of D. laureola in our study region.
Keywords: abiotic conditions, beetle pollination, canopy cover, Daphne laureola, distance to the nearest neighbour, gynodioecy,
hemispherical canopy photographs, microhabitat, winter flowering.
Résumé : Daphne laureola (Thymelaeaceae) est un arbuste gynodioïque à floraison hivernale qui habite le sous-étage de
forêts ombragées de montagne. Ses petites fleurs jaunâtre-vertes peu visibles sont pollinisées par des insectes ectothermiques
dont l’activité est principalement limitée aux rares périodes de temps ensoleillé. Nous avons évalué les hypothèses que (1)
l’environnement radiatif solaire des plants de D. laureola et la distance à l’individu conspécifique le plus proche devrait
influencer leur succès de pollinisation et (2) l’effet des caractéristiques spécifiques du site sur le succès de pollinisation
devrait dépendre du sexe, étant donné la différence entre les fleurs femelles et hermaphrodites quant à leur visibilité et à leur
récompense florale. Nous avons évalué le succès de pollinisation maternel (le nombre moyen de tubes polliniques/pistils et
la proportion de fleurs non fécondées), mesuré la distance à l’individu conspécifique en floraison le plus proche et évalué
l’environnement radiatif (le pourcentage de fermeture du couvert) pour des plants individuels de D. laureola dans 7 populations
du sud-est de l’Espagne. En moyenne, les fleurs hermaphrodites ont été pollinisées plus souvent et avaient plus de tubes
polliniques/pistils que les fleurs femelles. La distance à l’individu conspécifique le plus proche a eu un effet négatif sur
le succès de pollinisation individuel, mais cet effet n’était significatif que pour les femelles. De plus, l’interaction entre le
sexe et le couvert était fortement significative pour le succès de pollinisation. Chez les individus femelles, le nombre de
tubes polliniques/pistils a diminué et la proportion de fleurs non fécondées a augmenté des sites à couvert fermé à ceux plus
ouverts, alors que les hermaphrodites ont démontré des tendances opposées, mais non significatives. On s’attend donc à ce
que la relation inverse pour les 2 genres entre le couvert forestier et le succès de pollinisation crée une mosaïque à petite
échelle au sein de l’habitat pour ce qui est de l’ampleur de l’avantage pollinique des individus hermaphrodites, cet avantage
serait plus important dans les sites plus ouverts et ensoleillés où les pollinisateurs ectothermiques devraient être plus actifs.
Les différences spécifiques du site quant au succès de pollinisation selon le sexe pourraient contribuer à la persistance de la
gynodioécie chez D. laureola dans notre région d’étude.
Mots-clés : conditions abiotiques, couvert forestier, Daphne laureola, distance au voisin le plus proche, floraison hivernale,
gynodioécie, microhabitat, photographies hémisphériques de la canopée, pollinisation par des coléoptères.
Nomenclature: Castroviejo et al., 1997.
1Rec.
2007-06-18; acc. 2007-12-28.
Associate Editor: Johannes Kollmann.
2Author for correspondence.
DOI 10.2980/15-3-3115
ÉCOSCIENCE, vol. 15 (3), 2008
Introduction
In animal-pollinated plants, individual variation in
intrinsic flower and inflorescence features generally contributes to individual differences in pollination success and seed
output, in accordance with the decisive selective role played
by pollinators on these traits (Pellmyr, 2002). Nevertheless,
consideration of the ecological context (i.e., the extrinsic biotic and abiotic conditions) in which the interaction
of plants and pollinators takes place is also essential to a
proper understanding of its evolutionary potential (Herrera,
1996; Galen, 1999; Ashman, 2006). The ecological context
can affect the interaction of plants with their pollinating
agents through a variety of mechanisms. For example,
plants can respond to variations in the abiotic environment by plastically modifying certain features that may
eventually affect their attractiveness to pollinators, such as
floral display (Salonen, 1994; Schemske & Bierzychudek,
2001; Totland, 2001), flower size and shape (Murcia, 1990;
Herrera, 2005), nectar production (Petanidou & Smets,
1996), and floral longevity and phenology (Peterson, 1997;
Utelli & Roy, 2001). In addition, spatio-temporal variations in abiotic conditions may induce changes in pollinator
composition, abundance, and behaviour that will in turn
give rise to differences in floral visitation, pollen export
and receipt (Warren, Harper & Booth, 1988; Eckhart, 1992;
Herrera, 1995; Eckert, 2002; Alonso, 2005; Price et al.,
2005; Ashman, 2006), and even the sign and strength of
selection on floral traits (Caruso, Peterson & Ridley, 2003;
Herrera, Castellanos & Medrano, 2006). Finally, the biotic
environment, as represented by the diversity, local density,
and spatial distribution of simultaneously flowering plants,
may affect plant–animal interactions in general (see, e.g.,
Karban, 1997 and references therein for herbivory), particularly plant–pollinator interactions, through effects on insect
activity, foraging behaviour, and pollination effectiveness
(Laverty, 1992; Feldman, 2006; Metcalfe & Kunin, 2006).
Each of the preceding effects can take place at different
spatial scales, involving variations at the among- and within-plant population levels. In this paper we will focus on the
pollination consequences of small-scale spatial variations in
the ecological context, or “site-specific” effects, occurring
at the within-population level. This aspect of plant–pollinator interactions has been only infrequently examined (but
see Herrera, 1995; 1997; Orueta, 2002; Sánchez-Lafuente
et al., 2005), but it is important because site-specific effects
may limit the responses of plant populations to selection by
pollinators on intrinsic floral traits (Herrera, 1996).
Microclimate and the distribution of conspecific flowering neighbours are possibly the 2 most important determinants of site-specific pollination effects within plant
populations. Microclimatic effects on pollination success
should be most apparent in plants pollinated by ectothermic
insects that occupy habitats like the forest understory, characterized by patchiness in environmental variables potentially affecting pollinator foraging (Beattie, 1971; Herrera,
1995). The patterns of solar irradiance on a forest floor produce a changing mosaic of intensities, which provides a rich
and varying array of microhabitats for plants and insects
(Herrera, 1995; 1997). When flowers are larger than pollinators they may provide favourable environments in which
shelter and temperature can be experienced by pollinators
as a reward (Seymour, White & Gibernau, 2003; Sapir,
Schmida & Neeman, 2006). On the other hand, the spatial
distribution of conspecifics, and particularly the distance
between neighbours, may contribute to generate site-specific
effects on pollination success in species whose individuals depend strictly on pollen vectors for sexual reproduction (e.g., dioecious and self-incompatible hermaphroditic
species; Widén & Widén, 1990; Kunin, 1992; Metcalfe &
Kunin, 2006). Site-specific effects on pollination success
related to variation in microclimate and spatial relationships
to conspecifics are therefore expected to be most frequent
among sexually polymorphic species pollinated by ectothermic insects.
In this paper we report on a study designed to evaluate
the effects of the solar irradiance environment and distance
to the nearest conspecific neighbour on the maternal pollination success of female and hermaphrodite individuals
of Daphne laureola, a gynodioecious understory shrub.
In gynodioecious species, pollen limitation of seed production of female plants often increases with distance to
pollen donors (Widén & Widén, 1990; Graff, 1999; but
see Pettersson, 1997). However, we are not aware of any
previous study on the effects of abiotic heterogeneity in
sex-differential reproductive success in gynodioecious species (but see Barr, 2004 for an experimental approach). In
our study region (see below), D. laureola typically grows
in the understory of coniferous and mixed montane forests, all populations are gynodioecious, and female and
hermaphrodite individuals are similar in size and floral display (Alonso & Herrera, 2001). Individual female flowers
are less conspicuous because of shorter corolla tubes and
absence of pollen. Flowering starts during winter, and the
species is mainly pollinated by a small, ectothermic nitidulid pollen beetle (Alonso, 2004). We tested the hypotheses
that (1) the solar irradiance environment experienced by
individual D. laureola plants, and the distance to the nearest
neighbour, should influence their pollination success, and
(2) the effect of site-specific variables on pollination success should be sex-dependent, given the difference between
female and hermaphrodite flowers in the nature of floral
rewards, conspicuousness, and dependence on pollinators
for pollen receipt. We expected a stronger effect of isolation
on the pollination success of female plants, but we did not
have a priori expectations about the nature of gender-specific effects of the solar irradiance microenvironment.
Methods
Study species and sites
Daphne laureola (Thymelaeaceae) is an early-season
flowering, evergreen shrub growing in the undergrowth of
shady mountain forests. In the Iberian Peninsula, the species is frequent in the northern Cantabrian and Pyrenean
Mountains and also in the southern Betic Ranges but is
absent from the central Iberian Ranges (Alonso, Mutikainen
& Herrera, 2007). This study was conducted in 2002 in
7 different populations within the Natural Park of Sierras
de Cazorla, Segura y Las Villas (Jaén province, southeastern Spain). These populations are a subset of those
used in an earlier study on altitudinal gradients of pollination
359
Alonso & Herrera: Site-specific effects on pollination success
success (see Alonso, 2005 for further details) and encompass most of the species’ altitudinal range in the region
(see Table I). Distances between nearest sites range
between 3 and 22 km. All the study populations were
located in old, well-preserved natural forests. A mixture of
Pinus nigra, P. pinaster, and Quercus ilex dominated the
evergreen canopy in all study sites except Coto del Valle,
where the forest was largely dominated by Quercus faginea. Crataegus monogyna and Prunus mahaleb, 2 deciduous treelets, also occurred frequently in the overstory of
our study sites. All deciduous species usually start leafing
after D. laureola has completed flowering.
In the study region D. laureola populations consist of
a mixture of hermaphrodite and female individuals, with
female frequency being highly variable (3.8–64.3%; 94 populations) and decreasing with elevation (Alonso & Herrera,
2001). The small, tubular, yellowish-green flowers open synchronously in late winter and are aggregated into compact
axillary inflorescences. Pistils have a single ovule. Female
flowers have vestigial stamens that do not produce pollen
and have shorter corolla tubes (mean ± SD: 4.9 ± 0.45 mm,
n = 45 flowers) than the perfect flowers of hermaphrodites
(8.1 ± 0.92 mm, n = 56). Hermaphroditic plants are selfcompatible, but fruit production requires flower visitation
by pollinators (Alonso & Herrera, 2001). The pollen beetle
Meligethes elongatus (Coleoptera: Nitidulidae) and, to a
much smaller extent, small solitary bees and noctuid moths
are the main pollinators (Alonso, 2004).
Site-specific features of individual plants
Six female and 6 hermaphrodite D. laureola individuals
were haphazardly chosen in each of the 7 study populations.
Site-specific features of each plant were characterized by
the distance to the nearest flowering conspecific, measured
between the edges of the plants, and a quantitative estimate
of the forest canopy cover above it. We used compass-referenced hemispherical photographs of the forest canopy taken
around the midpoint of the flowering season to estimate
canopy cover above individual plants. These estimates were
then used to indirectly characterize the site-specific light
environment and, more specifically, the mean daily amount
of solar radiation received by each plant (Chazdon & Field,
1987; Becker, Erhart & Smith, 1989). The cumulative light
level reaching a location on the forest floor is a complex
function of the size, shape, and orientation of the portion
of the sky hemisphere that is visible from it (Canham,
1988; Yoshida, Yanagisawa & Kamitani, 1998). Although
hemispherical photographs may sometimes produce biased
estimates of canopy gap fractions (Bellow & Nair, 2003),
they are generally considered a robust indirect method
of estimating light transmittance to the forest understory
(Lhotka & Loewenstein, 2006), and we are confident that
the accuracy of this method was suitable for the purposes of
this study. Photographs were taken vertically from the top of
each plant using a 180° hemispherical (fisheye) lens (Nikon
FC-E8) and a digital camera (Nikon Coolpix 995) assembled on a tripod. The circular images depicting the canopy
above each plant were analyzed using the program WinPhot
5.0 (Steege, 1996) to obtain the percentage of canopy cover.
Two plants with defective hemispherical photographs were
excluded from the analyses, which finally comprised data
for n = 82 plants.
Pollination success
At the end of the flowering period, 12 flowers were collected from each study plant and preserved in FAA solution
(2.5% formaldehyde, 2.5% acetic acid, 95% ethanol) until
dissection and measurement. Pistils were softened in 1 N
KOH at 65 °C for 20 min, rinsed with distilled water, and
stained for 20 min at 65 °C in decolourized aniline blue. The
number of pollen tubes in each pistil were counted under an
epifluorescence microscope for a total of n = 984 flowers.
The maternal pollination success of individual plants was
then estimated using 2 different but complementary variables: the mean number of pollen tubes per pistil, which
provided an estimate of successful visits, and the proportion
of flowers within a plant without pollen tubes (unfertilized
flowers hereafter), which provided an index of the frequency
of pollination failure.
Data analyses
All statistical analyses were performed using the SAS
statistical package (SAS Institute, 2002). The effects of sex,
distance to nearest neighbour, percent canopy cover, and
the pairwise interactions sex × distance and sex × cover
on the mean number of pollen tubes per pistil and the proportion of unfertilized flowers per plant were analyzed by
fitting generalized linear mixed models to the data using
the GLIMMIX procedure (as implemented in SAS version 9.1.3; Littell et al., 1996), with Poisson and binomial
error distributions, respectively. The 2 models fitted the
data closely, as shown by normal distribution of residuals and generalized χ2 / degrees of freedom ratios close to
unity (1.03 and 1.42, respectively). The three-way interac-
Table I. Site elevation and descriptive statistics for the distance to the nearest flowering conspecific and canopy cover of Daphne laureola study
plants (n = 12 per site). Shown are the average values (SD in parentheses) and the 25% and 75% percentiles (P25 and P75, respectively).
Site
Elevation (m asl)
Coto del Valle
950
Cuevas Bermejas
1210
Roblehondo
1235
Prados de Navahondona
1490
Cañada de la Medianega
1560
Cañada del Espino
1575
Puerto Llano
1800
All combined
360
Distance (m)
Canopy cover (%)
Mean (SD)
P25
P75
Mean (SD)
P25
P75
1.3 (1.3)
1.7 (1.1)
2.9 (2.7)
0.6 (0.5)
1.0 (1.7)
0.3 (0.4)
1.2 (2.2)
2.0 (5.2)
0.7
1.1
1.1
0.1
0.0
0.0
0.0
0.0
1.4
2.9
4.0
1.0
5.5
0.8
0.5
1.6
72.0 (4.6)
82.0 (1.8)
79.5 (4.0)
76.2 (2.1)
71.7 (4.7)
78.0 (1.3)
68.3 (6.3)
75.7 (5.8)
67.2
81.6
77.1
75.3
69.9
77.1
62.4
71.9
76.3
83.0
82.3
77.5
75.3
78.6
74.5
79.5
ÉCOSCIENCE, vol. 15 (3), 2008
tion sex × distance × cover was not statistically significant
(P > 0.30) and was excluded from the final models. Since
our focus here was on individual variation within populations rather than on possible differences among populations,
the effect of populations was considered as random. Unless
otherwise stated, least-squares means (± SE) estimating the
marginal means corresponding to the fixed effects over a
balanced population are shown.
Results
The distance from focal plants to the nearest conspecific ranged between 0 and 42.4 m (all populations combined). Although this broad range was heavily influenced by
a few extreme data, among-plant variation in distance to the
nearest conspecific was still apparent when the interquartile
range, a more robust variation statistic defined by the 25 and
75% percentiles, was considered (Table I). As expected for
an understory species, percent canopy cover was relatively
high at all sites (Table I), but it was also variable between
individuals, ranging between 61 and 86% (all populations
combined). Distance to the nearest neighbour and canopy
cover were not significantly correlated across plants in our
sample (rs = 0.17, n = 82, P = 0.13).
Female and hermaphrodite plants did not differ significantly in either distance to the nearest conspecific
(χ21 = 0.92, P = 0.59; Kruskal–Wallis test) or percent canopy cover (χ21 = 0.02, P = 0.89). On average, and for all
populations combined, flowers of female plants were pollinated significantly less often than flowers of hermaphrodites
(proportion of unfertilized flowers/plant = 0.64 ± 0.04 and
0.14 ± 0.02 for females and hermaphrodites, respectively;
χ 21 = 51.4, P < 0.0001). Flowers of female plants also
had significantly fewer pollen tubes/pistil on average than
flowers of hermaphrodites (1.43 ± 0.26 and 14.16 ± 1.01
pollen tubes for females and hermaphrodites, respectively;
χ21 = 60.8, P < 0.0001).
The absence of differences between sexes in both sitespecific variables in our sample facilitated unbiased tests of
sex-dependent effects on pollination success. Results of the
generalized linear mixed models testing for the effects of
sex, canopy cover, distance to nearest conspecific, and their
interactions on the measurements of pollination success
are summarized in Table II. Of the 3 main effects tested,
plant sex was the only non-significant one, which contrasts
with the significant differences between sexes noted earlier,
obtained when the data are analyzed without simultaneously
considering individual differences in canopy cover and
distance to nearest neighbour. Percent canopy cover had a
barely significant effect on pollen tubes per flower but not
on the proportion of unfertilized flowers. Distance to the
nearest conspecific had highly significant effects on both
the number of pollen tubes and the proportion of unfertilized flowers. These results clearly demonstrate that the
2 site-specific plant features considered in this study are
important determinants of the pollination success of individual D. laureola plants. The sign and magnitude of the
effects of site-specific features on pollination success, however, are contingent on plant sex, as shown by the statistical significance of the 2 interactions included in the model
Table II. Results of the generalized linear mixed models testing for
the effect of sex, canopy cover, and distance to the nearest conspecific on 2 complementary measurements of pollination success of
Daphne laureola individuals. Statistically significant effects are
shown in bold type. The interaction S × C × D was far from significant and consequently excluded from the final models. Population
was included as a random effect in models.
Effect
Number of pollen
tubes per flower
F1, 70
P
Proportion of
unfertilized flowers
F1, 70
P
Sex (S)
Cover (C)
Distance (D)
S × C
S × D
0.03
4.68
7.33
9.16
6.12
0.21
0.75
10.52
5.20
6.21
0.863
0.034
0.009
0.004
0.016
0.647
0.391
0.002
0.026
0.015
(Table II). Similar results were obtained when the analyses
were conducted on the n = 54 plants for which the closest
neighbour was a potential pollen donor, i.e., a hermaphrodite plant (not shown).
The effects of plant canopy cover and distance to
nearest neighbour on individual pollination success were
different for female and hermaphrodite plants, as revealed
by statistically significant sex × cover and sex × distance
interactions in Table II. The nature of these interactions
is revealed by plotting the original data separately for the
2 sexes (Figure 1). Females and hermaphrodites differed in
the slope of the relationship linking pollination success and
percent canopy cover (Figure 1a,c). The pollination success – canopy cover relationships were not significant for
hermaphrodites (P ≥ 0.22; Figure 1a,c, open dots). In contrast, in the case of female plants there was a significant
direct relationship between pollen tubes per flower and
canopy cover (F1, 32 = 11.61, P = 0.002) and an inverse
relationship between the proportion of unfertilized flowers and canopy cover (F1, 32 = 6.35, P = 0.017). Plants of
both sexes experienced significant declines in pollination
success with increasing distance to the nearest conspecific,
but the effect was most marked for females (Figure 1d).
The general picture emerging from examination of the sex
× cover and sex × distance interactions, therefore, is that
the pollination success of female plants was much more
sensitive to small-scale variations in site-specific features
than that of hermaphrodites.
Discussion
The separation of sexual functions in plants may be
associated with physiological and ecological differences
between individuals of different sexes that could eventually
contribute to the evolution and maintenance of polymorphic
breeding systems (Bierzychudek & Eckhart, 1988; Freeman
et al., 1997; Dawson & Geber, 1999). Indeed, spatial segregation of sexes along environmental gradients has been
repeatedly reported for dioecious (Bierzychudek & Eckhart,
1988; Freeman et al., 1997) and, less frequently, subdioecious species (El-Keblawy & Freeman, 1999; Schultz,
2003). Although empirical evidence of the proximate mechanisms is scarce, spatial segregation of sexes would be
favoured, in theory, in populations with heterogeneous environments whenever sex-differential reproductive success
361
Alonso & Herrera: Site-specific effects on pollination success
Figure 1. Relationships between estimates of the maternal pollination
success and site-specific features of female (filled dots) and hermaphrodite (open dots) individuals of Daphne laureola. Upper row: variation
of the mean number of pollen tubes per pistil with (a) the plant’s canopy
cover and (b) distance to the nearest conspecific. Lower row: relationship
between the proportion of unfertilized flowers and the plant’s (c) canopy
cover and (d) distance to the nearest conspecific. Note that the regression
lines in panels b and d are plotted only for illustrative purposes because
zeroes in horizontal axis were not represented due to the logarithmic scale.
or sex-differential mortality occurs across environments
(Bierzychudek & Eckhart, 1988). This study has shown
that individual pollination success of females and hermaphrodites in the gynodioecious D. laureola was affected
differentially by the 2 site-specific features considered, a
finding suggesting that spatial segregation of sexes may
potentially occur in this species. In seeming contradiction
with this hypothesis, we found that differences between
sexes for both distance to the nearest conspecific and
canopy cover were not statistically significant in our study
sample. This result, however, does not necessarily rule out
the possibility of spatial segregation of sexes or sexual
selection of microhabitat occurring in D. laureola, because
our sampling design involved rather few randomly sampled individuals in each of several populations instead of
analyzing comparatively the site-specific characteristics
of random samples of plants of the 2 sexes within populations, which would have required much larger samples per
site. Elucidation of the potential consequences of observed
patterns in terms of spatial segregation of sexes, and their
possible correlates with local sex ratios (Graff, 1999), must
thus await future studies.
Distance-dependent fecundity of female plants related to
variation in pollination success has been previously reported
for both dioecious (de Jong, Batenburg & Klinkhamer,
2005) and gynodioecious insect-pollinated species (Widén
& Widén, 1990). In this study, individual pollination success of D. laureola was negatively affected by distance
362
to the nearest conspecific, and as expected, the effect was
stronger for females than for the self-compatible hermaphrodites of this winter-flowering gynodioecious plant. The
quantitative differences in pollination success reported here
are expected to be particularly significant for female individuals. Daphne laureola flowers have a single ovule, and
the mean number of pollen tubes per flower of female plants
(1.4) was only slightly above the minimum indispensable
for setting a fruit. Furthermore, the proportion of unfertilized flowers in females (which is negatively correlated with
fruit set; Alonso, 2005) ranged between 100 and 8%, which
likewise indicates a higher frequency of insufficient pollinations among females when compared to hermaphrodites, for
which the proportion of unfertilized flowers per plant was
always below 50% (Figure 1d). While variations in the distance to the nearest conspecific at the relatively small scale
of 0–10 m are consequential in terms of fitness for female
plants, the female function of hermaphrodite individuals was independent of distance to the nearest conspecific
(Figure 1b,d). This finding is congruent with the extremely
high selfing rates obtained in a subset of the populations
studied here (Medrano, Alonso & Herrera, 2005), which
would suggest a high frequency of pollinator-induced selfing and/or geitonogamous pollinations, and hence a relative
independence of pollinator movements, among hermaphrodite plants. However, the possibility that among-plant differences in spatial isolation may affect pollination quality
(e.g., proportions of self and cross pollen received) and the
outcrossing rates of hermaphrodites’ progenies cannot be
discarded (see, e.g., van Treuren et al., 1993; García et al.,
2005).
Solar irradiance and ambient temperature may influence the frequency and identity of floral visitors to individual plants through differences in pollinator thermal biology
and microhabitat selection (Herrera, 1995; 1997) and in the
pollinating efficiency of floral visits due to changes in stigma receptivity and pollen germination rates (Murcia, 1990;
Orueta, 2002; Galen & Stanton, 2003; Hedhly, Hormaza
& Herrero, 2005). Scarcity of information on the effect of
the solar irradiance environment on the pollination success
of individual understory plants (Herrera, 1995; SánchezLafuente et al., 2005), and the absence of such information
for gynodioecious species, prevented formulating any plausible a priori hypothesis on the possible differential effects
of canopy cover on the pollination success of plants of
the 2 sexes. In principle, pollination success of understory
hermaphroditic species flowering under adverse environmental conditions for pollinator activity is expected to be
affected by variations in microclimate (Herrera, 1995), and
thus both female and hermaphrodite plants were expected
to be affected. We found, however, a highly significant sex
× cover interaction effect on measurements of individual
pollination success, and this effect cannot be attributed to
potential artifacts due to covariation of cover with plant
distance since these 2 variables were uncorrelated in our
sample. Among female individuals, the number of pollen
tubes/pistil decreased and the proportion of unfertilized
flowers increased from closed to more open sites, while hermaphrodites showed non-significant trends. The contrasting
sign of the relationship between cover and pollination suc-
ÉCOSCIENCE, vol. 15 (3), 2008
cess for the 2 genders is expected to generate a fine-scale,
within-habitat mosaic in the magnitude of the pollination
advantage of hermaphrodite individuals, which should be
greater in more open, sunny locations where ectothermic
pollinators should be more active. Indeed, the most frequent
pollinator observed in our study region, the ectothermic
pollen-feeding beetle Meligethes elongatus, shows definite
preference for sunny locations (providing a favourable
thermal environment) and hermaphrodite flowers (providing pollen as food) (Alonso, 2004). Beetle ability to discriminate by visual and olfactory cues against the pollenless
individual females, and/or the residence time spent crawling
among their flowers (see Cook et al., 2002; Cook, Murray
& Williams, 2004 for M. aeneus), would decline in the more
closed locations. In relation to this interpretation, it is interesting to note that Sánchez-Lafuente et al. (2005) found no
effects of the irradiance environment on the pollination success of Helleborus foetidus, another understory winter-flowering species that coexists locally with D. laureola in our
study region. In contrast to beetle-pollinated D. laureola,
H. foetidus is pollinated by bumblebees, whose endothermic
capacity would make them independent of sunny patches
in the forest. Certainly, the influence of canopy cover on
D. laureola pollination success could vary geographically if
the main pollinator varies regionally, since different insect
species might be differentially affected by the solar irradiance environment (Herrera, 1997).
Although we did not measure it, the male component
of pollination success in hermaphrodite plants could be also
affected by site-specific effects dependent on the irradiance environment. Given the very high levels of inbreeding depression exhibited by selfed progenies in some of
our study populations (Medrano, Alonso & Herrera, 2005),
it may be predicted that siring success of hermaphrodites
would be improved with increasing canopy cover because
of the increased likelihood of pollen transfer to nearby
female plants under shadier locations. Both the site-specific
differences in gender relative maternal pollination success
and the contrasting consequences for male and female functions of hermaphrodites might contribute to the persistence
of gynodioecy of D. laureola in our study region.
It is well known that successional stage, treefalls, and
other microscale disturbances determine a gap-dynamics
in forests that can strongly influence the composition and
population dynamics of plant communities (Laska, 2001).
Although it has been less frequently studied from an individual-centred perspective, gap dynamics may also be consequential for the fitness of individual plants (Laska, 2001).
The results reported here highlight that the pollination success of an understory species may be highly dependent on
site-specific effects such as canopy cover and distance to
conspecifics, and consequently on any natural or anthropogenic disturbance. They also underline that such effects
could vary between male and female reproductive functions,
suggesting that more studies on within-population, site-specific effects on female and male functions and in species
with polymorphic breeding systems are required to better
understand the possible significance of the small-scale ecological context in breeding system evolution.
Acknowledgements
We thank M. García, A. Manzaneda, and N. Varo for lab and
field assistance, two anonymous referees for their comments on
the manuscript, and the Consejería de Medio Ambiente, Junta de
Andalucía for authorizing this work in Cazorla. C. Alonso also
acknowledges a Ramón y Cajal grant provided by the Spanish
Ministerio de Educación y Ciencia. The study was funded by
the Spanish Ministerio de Educación y Ciencia through research
projects BOS2003-02235 and CGL2006-01355/BOS and the
Consejería de Innovación, Ciencia y Empresa, Junta de Andalucía,
through Excellence Research Project RNM156-2005.
Literature cited
Alonso, C., 2004. Early blooming challenges: Extended flowering
season, diverse pollinator assemblage, and the reproductive
success of gynodioecious Daphne laureola. Annals of Botany,
93: 61–66.
Alonso, C., 2005. Pollination success across an elevation and sex
ratio gradient in gynodioecious Daphne laureola. American
Journal of Botany, 92: 1264–1269.
Alonso, C. & C. M. Herrera, 2001. Neither vegetative nor reproductive advantages account for high frequency of male-steriles
in southern Spanish gynodioecious Daphne laureola. American
Journal of Botany, 88: 1016–1024.
Alonso, C., P. Mutikainen & C. M. Herrera, 2007. Ecological context of breeding system variation: Sex, size and pollination in a
(predominantly) gynodioecious shrub. Annals of Botany, 100:
1547–1556.
Ashman, T. L., 2006. The evolution of separate sexes: A focus on
the ecological context. Pages 204–222 in L. D. Harder & S. C.
H. Barrett (eds.). Ecology and Evolution of Flowers. Oxford
University Press, Oxford.
Barr, C. M., 2004. Soil moisture and sex ratio in a plant with
nuclear-cytoplasmic sex inheritance. Proceedings of the Royal
Society of London, B, 271: 1935–1939.
Beattie, A. J., 1971. Itinerant pollinators in a forest. Madroño, 21:
120–124.
Becker, P., D. W. Erhart & A. P. Smith, 1989. Analysis of forest
light environments. Part I. Computerized estimation of solar
radiation from hemispherical canopy photographs. Agricultural
and Forest Meteorology, 44: 217–232.
Bellow, J. G. & P. K. R. Nair, 2003. Comparing common methods
for assessing understory light availability in shaded-perennial agroforestry systems. Agricultural and Forest Meteorology,
114: 197–211.
Bierzychudek, P. & V. Eckhart, 1988. Spatial segregation of the
sexes of dioecious plants. American Naturalist, 132: 34–43.
Canham, C. D., 1988. An index for understory light levels in and
around canopy gaps. Ecology, 69: 1634–1638.
Castroviejo, S., C. Aedo, M. Laínz, P. Montserrat, R. Morales, F.
Muñoz Garmendia, G. Nieto Felinier, J. Paiva (eds.), 1997.
Flora Iberica. Real Jardin Botánica, CSIC, Madrid, Spain.
Caruso, C. M., S. B. Peterson & C. E. Ridley, 2003. Natural selection of floral traits of Lobelia (Lobeliaceae): Spatial and temporal variation. American Journal of Botany, 90: 1333–1340.
Chazdon, R. L. & C. B. Field, 1987. Photographic estimation of
photosynthetically active radiation: Evaluation of a computerized technique. Oecologia, 73: 525–532.
Cook, S. M., D. A. Murray & I. H. Williams, 2004. Do pollen beetles need pollen? The effect of pollen on oviposition, survival,
and development of a flower-feeding herbivore. Ecological
Entomology, 29: 164–173.
363
Alonso & Herrera: Site-specific effects on pollination success
Cook, S. M., E. Bartlet, D. A. Murray & I. H. Williams, 2002. The
role of pollen odour in the attraction of pollen beetles to oilseed
rape flowers. Entomologia Experimentalis et Applicata, 104:
43–50.
Dawson, T. E. & M. A. Geber, 1999. Sexual dimorphism in physiology and morphology. Pages 175–215 in M. A. Geber, T. E.
Dawson & L. F. Delph (eds.). Gender and Sexual Dimorphism
in Flowering Plants. Springer-Verlag, Berlin.
De Jong, T. J., J. C. Batenburg & P. G. L. Klinkhamer, 2005.
Distance-dependent pollen limitation of seed set in some insectpollinated dioecious plants. Acta Oecologica, 28: 331–335.
Eckert, C. G., 2002. Effect of geographical variation in pollinator fauna on the mating system of Decodon verticillatus
(Lythraceae). International Journal of Plant Sciences, 163:
123–132.
Eckhart, V. M., 1992. Spatio-temporal variation in abundance
and variation in foraging behavior of the pollinators of gynodioecious Phacelia linearis (Hydrophyllaceae). Oikos, 64:
573–586.
El-Keblawy, A. & D. C. Freeman, 1999. Spatial segregation by
gender of the subdioecious shrub Thymelaea hirsuta in the
Egyptian desert. International Journal of Plant Sciences, 160:
341–350.
Feldman, T. S., 2006. Pollinator aggregative and functional
responses to flower density: Does pollinator response to patches
of plants accelerate at low densities? Oikos, 115: 128–140.
Freeman, D. C., J. Lovett Doust, A. El-Keblawy, K. J. Miglia &
E. D. McArthur, 1997. Sexual specialization and inbreeding
avoidance in the evolution of dioecy. Botanical Review, 63:
65–92.
Galen, C., 1999. Why do flowers vary? The functional ecology of
variation in flower size and form within natural plant populations. BioScience, 49: 631–640.
Galen, C. & M. L. Stanton, 2003. Sunny-side up: Flower heliotropism as a source of parental environmental effects on pollen
quality and performance in the snow buttercup, Ranunculus
adoneus (Ranunculaceae). American Journal of Botany, 90:
724–729.
García, C., J. M. Arroyo, J. A. Godoy & P. Jordano, 2005. Mating
patterns, pollen dispersal, and the ecological maternal neighbourhood in a Prunus mahaleb L. population. Molecular
Ecology, 14: 1821–1830.
Graff, A., 1999. Population sex structure and reproductive fitness
in gynodioecious Sidalcea malviflora malviflora (Malvaceae).
Evolution, 53: 1714–1722.
Hedhly, A., J. I. Hormaza & M. Herrero, 2005. The effect of temperature on pollen germination, pollen tube growth, and stigmatic receptivity in peach. Plant Biology, 7: 476–483.
Herrera, C. M., 1995. Microclimate and individual variation in pollinators: Flowering plants are more than their flowers. Ecology,
76: 1516–1524.
Herrera, C. M., 1996. Floral traits and plant adaptation to insect
pollinators: A devil’s advocate approach. Pages 65–87 in D.
G. Lloyd & S. C. H. Barrett (eds.). Floral Biology: Studies on
Floral Evolution in Animal-Pollinated Plants. Chapman & Hall,
New York, New York.
Herrera, C. M., 1997. Thermal biology and foraging responses of
insect pollinators to the forest floor irradiance mosaic. Oikos,
78: 601–611.
Herrera, C. M., M. C. Castellanos & M. Medrano, 2006.
Geographical context of floral evolution: Towards an improved
research programme in floral diversification. Pages 278–294 in
L. D. Harder & S. C. H. Barrett (eds.). Ecology and Evolution
of Flowers. Oxford University Press, Oxford.
364
Herrera, J., 2005. Flower size variation in Rosmarinus officinalis:
Individuals, populations and habitats. Annals of Botany, 95:
431–437.
Karban, R., 1997. Neighbourhood affects a plant’s risk of herbivory and subsequent success. Ecological Entomology, 22:
433–439.
Kunin, W. E., 1992. Density and reproductive success in wild
populations of Diplotaxis erucoides (Brassicaceae). Oecologia,
91: 129–133.
Laska, G., 2001. The disturbance and vegetation dynamics: A review
and an alternative framework. Plant Ecology, 157: 77–99.
Laverty, T. M., 1992. Plant interactions for pollinator visits: A test
of the magnet species effect. Oecologia, 89: 502–508.
Lhotka, J. M. & E. F. Loewenstein, 2006. Indirect measures for
characterizing light along a gradient of mixed-hardwood riparian forest canopy structures. Forest Ecology and Management,
226: 310–318.
Littell, R. C., G. A. Milliken, W. W. Stroup & R. D. Wolfinger,
1996. SAS System for Mixed Models. SAS Institute, Cary,
North Carolina.
Medrano, M., C. Alonso & C. M. Herrera, 2005. Mating system,
sex ratio, and persistence of females in the gynodioecious shrub
Daphne laureola L. (Thymelaeaceae). Heredity, 94: 37–43.
Metcalfe, D. B. & W. E. Kunin, 2006. The effects of plant density
upon pollination success, reproductive effort and fruit parasitism in Cistus ladanifer L. (Cistaceae). Plant Ecology, 185:
41–47.
Murcia, C., 1990. Effect of floral morphology and temperature on
pollen receipt and removal in Ipomoea trichocarpa. Ecology,
71: 1098–1109.
Orueta, D., 2002. Thermal relationships between Calendula arvensis inflorescences and Usia aurata bombyliid flies. Ecology,
83: 3073–3085.
Pellmyr, O., 2002. Pollination by animals. Pages 157–184 in C.
M. Herrera & O. Pellmyr (eds.). Plant–Animal Interactions. An
Evolutionary Approach. Blackwell, Oxford.
Petanidou, T. & E. Smets, 1996. Does temperature stress induce
nectar secretion in Mediterranean plants? New Phytologist,
133: 513–518.
Peterson, M. A., 1997. Host plant phenology and butterfly dispersal: Causes and consequences of uphill movement. Ecology,
78: 167–180.
Pettersson, M. W., 1997. Solitary plants do as well as clumped ones
in Silene uniflora (Caryophyllaceae). Ecography, 20: 375–382.
Price, M. V., N. M. Waser, R. E. Irwin, D. R. Campbell & A. K.
Brody, 2005. Temporal and spatial variation in pollination of a
montane herb: A seven-year study. Ecology, 86: 2106–2116.
Salonen, V., 1994. Growth and reproduction of Rubus saxatilis in
relation to light availability. Acta
�����������������������������
Oecologica, 15: 485–493.
Sánchez-Lafuente, A. M., J. Guitián, M. Medrano, C. M. Herrera,
P. J. Rey & X. Cerdá, 2005. Plant traits, environmental factors,
and pollinator visitation in winter-flowering Helleborus foetidus (Ranunculaceae). Annals of Botany, 96: 845–852.
Sapir, Y., A. Shmida & G. Neeman, 2006. Morning floral heat as
a reward to the pollinators of the Oncocyclus irises. Oecologia,
147: 53–59.
SAS Institute, 2002. SAS for Windows (version 9.1.3). SAS
Institute, Cary, North Carolina.
Schemske, D. W. & P. Bierzychudek, 2001. Evolution of flower
color in desert annual Linanthus parryae: Wright revisited.
Evolution, 55: 1269–1282.
Schultz, S. T., 2003. Sexual dimorphism in gynodioecious
Sidalcea hirtipes (Malvaceae). I. Seed, fruit and ecophysiology.
International Journal of Plant Sciences, 164: 165–173.
ÉCOSCIENCE, vol. 15 (3), 2008
Seymour, R. S., C. R. White & M. Gibernau, 2003. Heat reward
for insect pollinators. Nature, 426: 243–244.
Steege, H. T., 1996. WinPhot 5: A Programme to Analyze
Vegetation Indices, Light and Light Quality from Hemispherical
Photographs. Tropenbos Guyana Reports 95-2, Tropenbos
Guyana Programme, Georgetown, Guyana [Online] URL:
http://www.bio.uu.nl/~herba/Guyana/winphot/wp_index.htm
Totland, O., 2001. Environment-dependent pollen limitation and
selection on floral traits in an alpine species. Ecology, 82:
2233–2244.
Utelli, A.-B. & B. A. Roy, 2001. Causes and consequences of floral
damage in Aconitum lycoctonum at high and low elevations in
Switzerland. Oecologia, 127: 266–273.
Van Treuren, R., R. Bijlsma, N. J. Ouborg & W. Van Delden, 1993.
The effects of population size and plant density on outcrossing
rates in locally endangered Salvia pratensis. Evolution, 47:
1094–1104.
Warren, S. D., K. T. Harper & G. M. Booth, 1988. Elevational
distribution of insect pollinators. American Midland Naturalist,
120: 325–330.
Widén, B. & M. Widén, 1990. Pollen limitation and distance-dependent fecundity in females of the clonal gynodioecious herb
Glechoma hederacea (Lamiaceae). Oecologia, 83: 191–196.
Yoshida, T., Y. Yanagisawa & T. Kamitani, 1998. An empirical
model for predicting the gap light index in an even-aged oak
stand. Forest Ecology and Management, 109: 85–89.
365